Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Measuring Social Media Credibility:
A Study on a Measure of Blog Credibility
by
Minjeong Kang
S. I. Newhouse School of Public Communications
Syracuse University
Submitted to the Institute for Public Relations
For the 2009 Ketchum Excellence in Public Relations Research Award
2
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Abstract
As individual members of publics now have greater expectations to have substantial interactions
with organizations, especially by using the digital/social media, communication professionals
have been looking for ways to enhance these interactions. After reviewing previous literature,
this current study intends to show that credibility of social media is one of the key factors
initiating and further increasing public engagement and communication effectiveness in the
digital/social media. Among various social media, this study limits its scope to blogs, which have
been considered to be the most pertinent social medium used in the field of communication
management. Previous research on blog credibility has reported limited reliability of measures,
partly due to the use of general credibility measures that have been developed to assess
credibility of traditional news media. An effective way to measure blog credibility is hard to
achieve without a valid, reliable method to insure that measurement. This study has developed
and validated a 14-item measure of blog credibility and by the usage of focus groups and a
survey. This study also discusses the implications practical aspects of measuring blog credibility.
Key Words: Credibility, Blogs, Social Media, Blog Credibility
3
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Measuring Social Media Credibility: A Study on a Measure of Blog Credibility
As the “boundary spanner” that mediates an organization and its strategic publics (Cutlip,
Center, & Broom, 2000), the public relations function essentially deals with the credibility of the
organization and the communicated messages. Therefore, to manage organization-public
communication strategically, credibility management is the key. In communication research, the
credibility of the communicator has widely been suggested to influence the processing of the
communicated content and the change of audience attitudes and beliefs (Burgoon & Hale, 1984;
Hovland, Janis, & Kelley, 1953). Recently, it has been suggested that the credibility of the
channel/medium of communication influences the selective involvement of the audience with the
medium (Metzger et al., 2003). Accordingly, individual audiences are paying closer attention to
the media that they perceive to be credible. When individual audiences rely more on a certain
communication medium for information seeking, they are likely to rate the medium more
credible than other media (Johnson & Kaye, 1998; Kiousis, 2001).
Fast-developing technology, especially in the digital media environment, has empowered
individuals and other stakeholder organizations to be creators of communication messages rather
than remaining as the static receiver of communication content (McClure, 2007, February 26).
As a result, unlike the old days when organizations were often the only senders of messages, the
communication process has increasingly become multidirectional, amplifying horizontal
influences among individual public members themselves. Additionally, as the influence of
emergent social media continues to increase, individual members of publics are starting to
demand participatory communication with and from organizations (Edelman, 2008, October 30).
4
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
In this era of social media, the demand for a public relations function based on direct and
trust-based communication with publics for better relationship building is great (Breakenridge,
2008). At the same time, the emergence of social media can also provide one of the best venues
for new public relations (Breakenridge, 2008; Scott, 2007; Weil, 2006), which can function as
the hub of dialogic, participatory communications between publics and organizations. In either
case, garnering credibility from publics is essential.
Instead of being passive recipients of messages, publics are now active enough to select
communication channels, especially when they care about an issue (Rubin, 2002). Unlike
traditional media such as radio or television, the use of social media by individuals is
increasingly becoming “instrumental” (Rubin, 1984), depending on their motivations for
communication (functionality), purposive or planned nature of communicating (intentionality),
communication choice (selectivity), and involvement with media. Given such audience selective
and active media use, communication without the creation and cultivation of credibility for better
public engagement or brand-building can be hardly effective (Holtz & Havens, 2009).
Among various venues of social media, the researcher considers blogs to be most
pertinent for the focus of this study. Blogs are driving “PR 2.0.”, as making a clear departure
from previous practice into much more interactive, spontaneous communication with publics
(Scoble & Israel, 2006; Weil, 2006; Yang & Lim, 2009). According to a recent survey of public
relations departments in the United States, 49 percent of respondents indicated their public
relations departments have been using blogs for practice (PR News, 2009, August 17). Therefore,
this current study will be strategically delimited to study a measure of blog credibility.
5
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
However, despite the plethora of discussion on social media credibility in the
blogosphere (Rowse, 2006, April 24), virtually no empirical research exists to validate the
measurement system of social media credibility. With the emergence of “new” audiences and
“new” media, the concept of credibility should be updated to the communication needs and
motives of new audiences as well as new functions of new media. Traditionally, in
communication research, the credibility of the communicator has been highlighted by
trustworthiness and expertise. In the contemporary blogosphere, however, blogger credibility has
often been replaced with emergent terms, such as “authenticity,” “legitimacy,” “transparency,”
“authority,” or “passion.” For example, the level of authenticity in the communicated messages
now decides the blogger’s credibility, rather than the communicator himself/herself (Banks, 2008;
Weil, 2006). Additionally, the legitimacy of the blogger is enhanced by the personal passion and
devotion to the communicated content bring out the legitimacy of the blogger, especially when
the blogger has authority about the knowledge to audiences who care (Scoble & Israel, 2006).
Without valid and reliable measurement of credibility, the management of credibility in
social media will be hardly feasible. Therefore, this current research aims to develop a new
measure of blog credibility from the perspective of new audiences, using exploratory focus
groups followed by a confirmatory survey.
LITERATURE REVIEW
The assessment of credibility in the online environment is often much more complex than
in previous media contexts due to “the multiplicity of sources embedded in the numerous layers
of online dissemination of content” (Sundar, 2008). Credibility has been discussed in the three
perspectives of communication: medium credibility, message/content credibility, and source
6
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
credibility (Metzger et al., 2003). Medium credibility is the perceived level of credibility that
individual users have of a specific medium, such as newspapers, television, the Internet, or blogs
(Newhagen & Nass, 1989; Sundar & Nass, 2001). Message credibility is the perceived
credibility of the communicated message itself, such as informational quality, accuracy, or
currency (Metzger et al., 2003). Past research on source credibility has focused on the expertise
or trustworthiness of the source as the likelihood to provide credible information (Armstrong &
Nelson, 2005; Berlo, Lemert, & Mertz, 1969; Burgoon & Hale, 1984; Flanagin & Metzger, 2003;
Hovland, et al., 1953; McCroskey & Richmond, 1996).
Past research on blog credibility has focused on credibility of blog medium compared
with traditional media and the Internet. Researchers used a general measure of medium
credibility across different media, generally used to measure traditional news media (Banning &
Sweetser, 2007; Johnson & Kaye, 1998, 2009; Thorson, Vraga, & Ekdale, 2010). For example,
Thorson et al. (2010) measured blog credibility into a single dimension using a scale comprised
of six semantic differentials: fair/unfair, biased/unbiased, accurate/inaccurate, doesn’t tell the
whole story/tells the whole story; cannot be trusted/can be trusted, and balanced/imbalanced,
which was developed by Meyer (1988) to evaluate credibility of news papers. Banning and
Sweetser (2007) used a general media credibility measure to assess blog credibility: factual;
concerned about making profits; invading people’s privacy; concerned about the community’s
well-being; and trusted. Johnson and Kaye (2009) also used a similar measure for credibility of
traditional media, comprised of believability, fairness, accuracy and depth.
7
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
As a result of using general credulity measures of traditional news media, previous
research on blog credibility reported limited measurement reliability. For example, Banning et al.
(2007) noted: “Cronbach alphas were run on the five credibility questions for the each of the
articles in the instrument, with the result that the Cronbach alpha for the first article was .65
while the Cronbach alpha for the second article was .54. These scores were consistent with those
reported in previous studies.” (p. 460). Given that the acceptable reliability cutoff point is often
at least greater than .80 (Tabachnick & Fidell, 2001), the researcher noted the need to develop a
blog-specific measure of credibility based on a Grounded Theory approach and coming from
blog users’ perspectives. The current research did not examine blog medium credibility, as there
are several previous studies on the topic (Banning & Sweetser, 2007; Johnson & Kaye, 1998,
2009; Thorson, et al., 2010). The purpose of this study is to examine the credibility within the
realm of blogs rather than across different media. Thus, the focus of the current study is to
develop and test a measure of blog credibility, delimiting its scope to blog source credibility and
blog message/content credibility.
Blog Medium Credibility
Blogs have often been doubted as a credible communication medium, as there is no clear
or “legitimate” gatekeeping process for content production and information sharing. Nonetheless,
it is important to understand credibility judgments from blog users’ perspectives, as credibility
judgments of blogs depend on who is being studied (Johnson & Kaye, 2009). When individual
users rely more on a specific medium for information, they consider it to be more credible than
other media (Flanagin & Metzger, 2000; Johnson & Kaye, 1998; Kiousis, 2001). Therefore,
8
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
individual blog readers who rely on blogs for information are likely to consider blogs as more
credible sources than other media.
The peer review process in the blogosphere can lead to the increase in the medium
credibility of blogs with blog readers considering blogs as moderate to very credible (Johnson &
Kaye, 2004). Among actual blog readers, Banning and Sweetser (2007) found that blogs were
rated more credible than any other online news source (e.g., online cable TV news, online news
magazines, online radio news) and even over traditional news sources. Experienced Internet
users rated blogs as more credible, due to their familiarity with the purpose of blogs and the style
of blog writing, (Hostway, 2005), and the fact they do not follow traditional news values such as
fairness, balance, and objectivity (Johnson & Kaye, 2009). Furthermore, Internet users consider
blogs to be credible because blogs are often independent from mainstream, corporate-controlled
media (Andrews, 2003), which can allow bloggers to write in-depth, opinionated messages in a
transparent manner (Scoble & Israel, 2006). Such transparency is the key factor driving blog
readers credibility judgments and audience engagement in blog-mediated messages (Yang &
Lim, 2009).
Blog Source Credibility: Blogger Credibility
Previous studies have examined blog source credibility. However, most of previous
research relied upon existing source credibility measures of traditional news media or the
Internet, without considering blog users’ perspectives. The problem with these previous
approaches is that without understanding how and why people find information in blogs more
trustworthy than other information sources, it is almost misleading to measure the level of
9
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
credibility using traditional credibility measures. As previously noted, it is critical to reflect blog
users’ perspectives in examining how credibility judgments are made.
Yang and Lim (2009) noted that individual users tend to trust organizations when they
perceive greater level of interactivity in social media; credibility of social media is critical in
enhancing such interactivity. Yang et al. found in their experimental research of blog-mediated
public relations that interactivity was strongly associated with trust. However, no significant
effect of blogger credibility was found. In order to measure blogger credibility, the authors used
a traditional measure of source credibility in relational communication (Burgoon & Hale, 1984):
trustworthy, expert, reliable, intelligent, professional and experienced. Yang, Kang, and Johnson
(2010) studied the effect of blog-mediated communication in mitigating individual stakeholders’
negative emotions and used the same 6-item measure of blogger credibility. Again, Yang et al.
did not find any significant effects of blogger credibility on key variables of the study, such as
audience engagement with crisis messages and their reduction of negative emotions. Therefore,
Yang et al. called for a Grounded Theory approach (Glaser & Strauss, 1967) from blog users’
perspectives to develop a valid measure of blogger credibility.
Blog Content Credibility
Minimal empirical research on blog content credibility exists currently. On the
conceptual level, a limited number books and blog posts have highlighted key attributes that
increase blog readers’ credibility judgments about blog content. For example, Banks (2008)
interviewed 30 leading bloggers and suggested the following attributes regarding blog content
credibility: credible blog contents are focused (i.e., delving into a specific time; establishing the
niche of personal passion), authentic (i.e., exclusive coverage of an interesting topic) and
10
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
insightful (i.e., in-depth opinions and rich personal experience). Rowse (2006, April 24) has
added consistency:
As I analyze which bloggers I see as credibly in the fields that I’m interested in one of the
factors that I see in all of them is a level of consistency… they don’t contradict
themselves in what they present, they are not swayed by popular opinion of them and
they produce quality content regularly over a long period of time. (para. 10)
Additionally, timeliness of blog content—i.e., frequent updates of content in the focused area—
has been suggested as a key trait of blog credibility (Banks, 2008; Weil, 2006).
METHODOLOGY
Preliminary Focus Groups
In order to explore key attributes of blog credibility, the researcher developed focus
groups with blog users in November 2009. Each of the five focus groups was comprised of 7 to
10 participants, with 41 blog users participating in total. According to Lenhart and Fox (2006), in
early 2006 from Pew Internet & American Life Project, more than half (approximately 54
percent) of bloggers were under the age of 30. Therefore, key criteria of selecting focus group
participants were experience of using/reading blogs on a regular basis, among individuals in their
20s to 30s, in conducting the single-category design of focus groups. The average age of focus
group participants was 25; more female participants (about 68 percent) attended focus groups.
As a pre-session strategy, a nominal group technique was used before focus group
sessions were held: Focus group participants were asked to make a list of key attributes of blog
credibility in general and then to rate the importance of each listed attribute. After the pre-session
11
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
nominal group technique, for focus group sessions were semi-structured with a questioning route
comprised of one opening question, one introductory question, and four key questions about
credibility judgments in blogger source and blog contents; and one ending question (i.e., “If you
have a chance to give advice to companies, what advice would you give them to improve their
communication credibility on social media?”). Focus group sessions lasted an average of one
hour. An experienced blog/social media user was selected as a moderator for the focus group and
trained in advance about the purpose of the research, focus group questioning route and other
moderating skills. Snacks and meals were offered during the session to stimulate conversation.
After each focus group, abridged transcripts were constructed for data analysis.
Several participants indicated that the source of blogs is critical. Accordingly, they made
judgments about source credibility “at first” to determine if they would read the blog or not. The
participants, who often prefer reading blogs over traditional news media, generally indicated that
blogs are a credible medium than other participants. Medium credibility of blogs was assessed
within other types of social media. In general, the order of social media credibility places blogs
higher than Twitter and Facebook, although it still depends on the kind of information users want.
Regardless of the different types of social media, focus group participants indicated that
companies need to be more interactive. Participants indicated a need for companies to be more
personal and approachable on social media sites in order to put a face to a faceless company, as
this serves as the basis of their trust toward companies.
The participants who write for blogs or have other substantial familiarity with blogs
suggested that credible bloggers are passionate, reliable and transparent, rather than applying
traditional source credibility attributes such as perceived expertise on the basis of blogger’s
12
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
educational or professional backgrounds. For example, some participants noted that they do not
trust such expertise information because people could lie in their bios on social media and say
they have expertise or degrees. Additionally, other participants also shared that their credibility
judgments of bloggers are often derived from the extent of the bloggers’ knowledge and
influence in the specific field of blog topics. Several participants indicated that they do not read
any blog on a regular basis unless bloggers are either personally known or leading/influential
individuals, or people with authority in the fields of their personal interests.
In regards to blog content credibility, strongly shared opinions existed among focus
group participants that credible blog contents are consistent and timely delivered of authentic,
insightful, and informative contents. This becomes a competitive advantage or the niche from
other traditional news media. When blogs covered the focused issue with dedicated efforts (e.g.
frequent updates, longevity of blogs, or interactive features and professional design of blogs),
participants also indicated that they found the content of these blogs as informative and insightful
as well. Among those who do not heavily rely upon blogs for information search and sharing,
common journalism criteria for information credibility were still applied for blog content
credibility: accuracy and fairness of information are still critical in credibility judgments of blogs.
“It’s the same as traditional journalism. There will always be things that you find out eventually
weren’t true. So time will always tell the truth. Inaccurate and biased information hurt blogs’
credibility significantly,” responded a participant, who said she often checks where the facts or
statistics come from.
On the basis of preliminary focus groups and literature review, the researcher developed a
list of key attributes in blogger source credibility and blog content credibility. As previously
13
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
noted, blog medium credibility was not examined on the survey, as the focus of the study was
not to compare blogs’ credibility with other media.
Survey Instrument
Following the results of exploratory focus groups and existing literature, the researcher
developed a measure to gauge blog credibility. The key purpose of this survey was to confirm
reliability of the proposed measure of blog credibility. Measurement instrument describes the
extent to which the respondents recognize the importance of each of the following 14 attributes
in evaluating credibility of blogs that they have been reading, in terms of two theoretical
dimensions (i.e., blogger credibility and blog content credibility), using 7-point Likert-type scale:
1. Blogger (source) credibility: knowledgeable; influential; passionate; transparent;
and reliable (5-item measure)
2. Blog message/content credibility: authentic; insightful; informative; consistent;
fair; focused; accurate; timely; and popular (9-item measure)
Survey Administration
In April 2010, a survey was conducted with individual users of social media for one week.
Participants completed a self-administered online questionnaire. A sample of 278 social media
users was taken using network sampling technique. The researcher posted a message to Twitter
and Facebook with a link to the online questionnaire asking social media users to participate in
the current study.
Demographic Profiles
14
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Demographic information was collected in terms of age (i.e., 18 to 25 = 45 percent; 26 to
30 = 31 percent; 31 to 45 = 20 percent; Over 45 = 4 percent), gender (i.e., female = 68 percent),
educational level (i.e., Mode is 4-Year college degree = 32 percent), and income level (i.e., Mode
is $50,000 to $74,999 = 41.5 percent). Among survey participants, 33 percent of them indicated
that they read blogs every day, followed by “a couple of times a week” (29 percent), “rarely,
only if I come across one randomly” (29 percent), and “once a week” (9 percent). About one
percent of sampled participants (n = 28) indicated that they never read blogs, so the researcher
did not analyze the results of their questionnaire. Therefore, the final sample size was adjusted to
n = 250.
Statistical Procedures for Data Analysis
AMOS 18.0 with Maximum Likelihood (ML) estimation was used as the statistical
package to analyze the data for hypothesized model. For scale purification and verification, the
researcher used both exploratory and confirmatory factor analyses. Exploratory factor analysis
(EFA) establishes a preliminary test of blog credibility measures by identifying items with low
factor loadings and determining whether each measurement item loaded on its intended factor
suggested by theories (Tabachnick & Fidell, 2001). Therefore, the primary goal of this step is to
assess (a) the dimensionality of measurement items into two theoretical factors and (b) the sound
association of each measurement item in their intended factor (i.e., standardized factor loading
greater than .65)(Hair, Anderson, Tatham, & Black, 1998).
Next, confirmatory factor analysis (CFA) examines and confirms the overall validity of
the measure the theoretical structure (Byrne, 2001). First, in order to establish discriminate
validity (i.e., difference between factors or latent variables), the researcher will compare the
15
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
goodness of data-model fits across different models with variant factor structures (i.e., in this
case, one-dimensional model and the proposed two-dimensional model). After establishing the
proposed two-dimensional model can be retained as a more reliable model, the researcher will
examine construct validity (i.e., reliability between items or indicators) of the proposed
measurement model.
Data-Model Fit Indexes
The study used four goodness-of-fit indexes to evaluate model fit: χ2
/df, the Comparative
Fit Index (CFI), the Root Mean Square Error of Approximation (RMSEA), and the SRMR
(Standardized Root Mean Square Residual). As a parsimonious index, χ2 /df needs to be less than
3 for a sound model fit (Byrne, 2001; Kline, 1998). Furthermore, aaccording to Hu and Bentler’s
(1999) joint-cutoff criteria for fit indexes in structural equation model (SEM), if the SEM model
has CFI ≥ .96 and SRMR ≤ 1.0 or RMSEA ≤ .06 and SRMR ≤ .10, it can be suggested that the fit
between the data and the proposed model is retainable as a valid model.
Data Screening and Mediation Test: Multivariate Normality and Bootstrapping
Before the analysis, the assumption of multivariate normality was tested. A common
procedure to address multivariate nonnormality is bootstrapping (Byrne, 2001). This study
performed a bootstrap on 1000 samples, using the ML estimates with bias-corrected 95%
confidence intervals for each of the bootstrap estimates. The bootstrap results did not deviate
from those based on normal theory; none of the bias-corrected 95% confidence intervals for
significant paths in the hypothesized models included zero, suggesting that the significant paths
indeed remained significant in the bootstrapping procedure.
16
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
RESULTS
Exploratory Factor Analysis
The researcher employed exploratory factor analysis in order to establish the
dimensionality of the data and sound association between measurement items in their intended
factors. As Table 1 indicates, the results supported the two-dimensional structure of blog
credibility measure into blogger credibility and blog content credibility. Also, Figure 1 illustrates
that there are clearly two discriminant groups of measurement items into the proposed structure.
Descriptive Statistics with Measurement Reliability
Blogger credibility. This dimension was used to measure the importance of each
attribute in the respondents’ answers making judgment of blogger credibility, using 7-point
Likert-type scale (i.e., 1 = not important at all; 4 = neutral; 7 = very important): (a)
knowledgeable (M = 6.28; SD = .67), (b) influential (M = 5.79; SD = .64), (c) passionate (M =
6.01; SD = .64), (d) transparent (M = 6.29; SD = .71), and (e) reliable (M = 6.11; SD = .72).
These 5 items loaded on one factor suggesting this measure is uni-dimensional, which explained
about 60.80 percent of the total variance. With regard to reliability, the resulting measure led to a
Cronbach’s alpha of .84.
Blog content credibility. Next, this dimension was used to measure the importance of
each attribute in the respondents’ making judgment of blog content credibility, using 7-point
Likert-type scale (i.e., 1 = not important at all; 4 = neutral; 7 = very important): (a) authentic (M
= 5.81; SD = .80), (b) insightful (M = 5.91; SD = .78), (c) informative (M = 5.94; SD = .75), (d)
consistent (M = 5.63; SD = .80), (e) fair (M = 5.62; SD = .73), (f) focused (M = 5.65; SD = .71);
17
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
(g) accurate (M = 6.06; SD = .76); (h) timely (M = 5.82; SD = .77); and (i) popular (M = 5.79; SD
= .70). These 9 items loaded on one factor suggesting this measure is uni-dimensional, which
explained about 65.46 percent of the total variance. With regard to reliability, the resulting
measure led to a Cronbach’s alpha of .93. Table 2 reports correlations between measurement
items.
Confirmatory Factor Analysis
Following Byrne (2001), the CFA model was slightly modified by allowing the following
error correlations among the observed items within the same subscale: e6-e7, e6-e8, and e9-e10
on the basis of data modification procedure. To the researcher, such error correlations are
theoretical. “Authentic” (e6) is positively correlated with “insightful” (e7) and “informative” (e8),
while “consistent” (e9) is positively associated with “fair” (e10).
To examine discriminant validity of two factors (i.e., “blogger credibility” and “blog
content credibility”), the researcher compared goodness of data-model fits across different
models with variant factor structures (i.e., in this case, one-dimensional model and the proposed
two-dimensional model).
First, the uni-dimensional model did not have sound data-model fits: 2
= 250.52, df = 74,
p < .001, χ2
/df = 3.93, CFI = .854, RMSEA = .128 (90% CI: .113, .143), and SRMR = .096. The
lowest standardized factor loading was .44 (“passionate”), while all loadings in were significant
at p < .001.
However, when applying two-dimensional model (see Figure 2), it turns out that data-
model fits of the CFA model was satisfactory enough to be retained as a valid model: 2
= 113.77,
18
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
df = 73, p < .001, χ2
/df = 1.56, CFI = .973, RMSEA = .056 (90% CI: .035, .075), and SRMR
= .042. The lowest standardized factor loading was .62 (“passionate”), while all loadings in were
significant at p < .001. In terms of 2-df tests, this two-dimensional model is suggested to have
sound discriminant validity of the two factors. This model with two factors performs much better
than the earlier model with one factor structure: 2
change (1, n = 250) = 136.75, p < .001. Also, all
data-model fit indexes became significantly better in the two-dimensional model: CFIchange
= .119, RMSEAchange = -.072, and SRMRchange = -.054.
In the final CFA model with two factors (see Figure 2), all factor loadings are greater
than .60 and significant at the .001 level, suggesting that each of the two factor has sound
construct validity. In terms of AVE (average variance extracted), the factor of blogger credibility
has .512, while the factor of blog content credibility has .602. This suggests that both factors
extract more common variances to explain the theoretical concept than error variances.
The researcher noted, from the final measurement model, that authority (to be
“influential”) and reliability of the blogger are highlighted in the blogger credibility factor, and
that accuracy and focus (as “the niche” of personal interests and experience) are key indicators
of the blog content credibility. Also, according to a factor plot (see Figure 1), the researcher
found that authority and reliability are very proximate, suggesting these two attributes are
strongly associated with each other. For example, a blogger with authority can be perceived to be
more reliable than other bloggers. Among indicators of the other factor (i.e., blog content
credibility), Figure 1 illustrates that three attributes (“authentic”, “insightful” and “informative”)
are proximate indicators, while “focused”, “consistent”, and “fair” are strongly associated.
19
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
DISCUSSION
The purpose of this research was to develop and test a measure of blog credibility in
multiple dimensions with blog users’ perspectives reflected. Several previous studies on blog
credibility were conducted to compare credibility of blogs with other traditional news media
(Banning & Sweetser, 2007; Johnson & Kaye, 1998, 2009; Thorson, et al., 2010). For the
purpose of comparative analysis, past research used general measures of medium credibility to
test traditional news media, such as Meyer’s (1988) measure of media credibility. Other
researchers (Yang, et al., 2010; Yang & Lim, 2009), also used exiting measures of
communication source credibility, such as in relational communication (Burgoon & Hale, 1984).
Previous research cited had limited reliability of the measure or failed to find significant effect of
blog credibility despite clear theoretical reasons. Therefore, researchers (e.g., Yang et al, 2010)
called for a more grounded approach, emergent from blog users’ perspectives, to develop and
test a measure of blog credibility. This current study was aimed to fill a void in previous research
on blog credibility.
Why is it important to develop a sound measure of blog credibility for the practice and
study of public relations or communication management? Communication professionals have
been facing various challenges in this new era of digital/social media to find ways to effectively
engage stakeholders that may lead to organizational support. As individual stakeholders expect
to have substantial interactions with other stakeholders, as well as with organizations,
organizations are also increasingly demanded to offer venues for two-way communication
allowing anyone to participate. Consequently, the increased interactions among stakeholders and
organizations can provide a sound basis to build relational trust, which can be a crucial
20
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
antecedent that leads to higher engagement by publics. Nonetheless, the missing link still
pertains as how these organizations can bring individuals to various social media that
organizations use for relationship management. In other words, what are key attributes of
successful social media that initiate interaction/relational loops with individual stakeholders?
Past studies suggest that credibility is the key. Instead of being a passive recipient of
messages, individual members of publics are now active enough to select communication
channels, especially when they care about an issue (Rubin, 2002). When individuals consider a
medium to be more credible than other media, they are also more likely to rely on that medium
for information search and sharing than other media (Johnson & Kaye, 1998; Kiousis, 2001).
Therefore, credibility can be suggested as one of the key factors driving the traffic of individuals
to organizations’ social media. Management of social media credibility can also be considered
the priority for enhancing effectiveness of the entire communication management process.
Furthermore, without a valid, reliable measure of social media credibility, successful
management of social media credibility is hardly feasible.
Among various social media, this study delimited its scope to blogs. As previously noted,
blogs are one of most commonly used social media by companies; blogs are driving “PR 2.0.”,
as making a clear departure from previous practice into much more interactive, spontaneous
communication with publics (Scoble & Israel, 2006; Weil, 2006; Yang & Lim, 2009).
This study conducted preliminary focus groups consistent with mainstream demographic
information of blog users (Lenhart & Fox, 2006). Focus group participants indicated blog-
specific attributes of credibility such as passionate, transparent, influential, authentic, insightful,
or focused, while emphasizing common credibility criteria such as knowledgeable, accurate, fair,
21
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
or consistent. This suggests that if a general measure of credibility is used to measure blog
credibility, important aspects of blog credibility are subject to being ignored. With triangulating
results of focus groups from existing literature on credibility, the researcher developed and tested
a 14-item measure of blog credibility.
The results of this study are very supportive of the proposed 14-item measure of blog
credibility. Reliability of this measure has been greatly improved, compared with previous
research. The data supports the two theoretical dimensions of blog credibility into blogger
(source) credibility and blog message/content credibility. Furthermore, the proposed
measurement model was successfully validated through confirmatory factor analysis in terms of
sound fits between the data and the model. According to key results of this study, authority (to
be “influential”) and reliability of the blogger are highlighted in the blogger credibility factor,
and that accuracy and focus (as “the niche” of personal interests and experience) are key
indicators of the blog content credibility. The results are consistent with previous literature on
blog credibility (Banks, 2008; Scoble & Israel, 2006; Weil, 2006). In terms of relations among
blog credibility indicators, the researcher also found that authority and reliability are very
proximate, suggesting these two attributes are strongly associated with each other. Among
indicators of the other factor (i.e., blog content credibility), the results of this study indicate that
three attributes (“authentic”, “insightful” and “informative”) are proximate indicators, while
“focused”, “consistent”, and “fair” are strongly associated.
Following this current study, the proposed measure of blog credibility can be applied to
investigate attitudinal and behavioral effects of blog credibility. For example, after assessing
blog credibility by employing content analysis or a survey with the proposed measure, future
22
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
research can examine how blog credibility brings out measurable effects such as blog traffic,
degrees of interactions, or blog readers’ positive attitudes and supportive behaviors. Past research
has suggested that enhanced credibility of a medium is strongly associated with increased
reliance on the medium for information and further behavioral outcomes. Future research can
conduct empirical tests about the link between blog credibility and attitudinal/behavioral
outcomes supportive to companies’ brands or reputations.
Future research can also study the role of blog design or aesthetic aspects of blogs in
terms of additional dimension of blog credibility. Several survey respondents indicated in their
open-ended responses that aesthetic aspects of blogs or professional blog design often become a
useful cue in deciding their first impressions of blogs and the basis of their credibility judgments.
This response was also indicated by focus group findings. Therefore, future researchers can
develop and test this additional dimension of blog credibility as blog design.
There are several limitations of the findings of this study. As the scope of this study was
kept to corporate communications and the general usefulness and credibility of information that
individuals find in blogs, the nature of information was restricted to contents about corporations
and their activities. As often is the case, the purpose and the credibility of information tend to co-
influence as the utility of information tends to affect the degree of efforts individuals exert to
discern the quality of information. This indicates that depending upon the purpose for
information, as previously noted, individuals’ credibility judgments also tend to shift. Individual
perceptions of credible information when he/she reads gaming blogs differ than when he/she
reads political blogs.
23
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
References
Andrews, P. (2003). Is blogging journalism? : Nieman Reports.
Armstrong, C. L., & Nelson, M. R. (2005). HOW NEWSPAPER SOURCES TRIGGER
GENDER STEREOTYPES. [Article]. Journalism & Mass Communication Quarterly,
82(4), 820-837.
Banks, M. A. (2008). Blogging heores: Interviews with 30 of the World's Top Bloggers.
Indianapolis, IN: Wiley Publishing, Inc.
Banning, S. A., & Sweetser, K. D. (2007). How Much Do They Think It Affects Them and
Whom Do They Believe?: Comparing the Third-Person Effect and Credibility of Blogs
and Traditional Media. [Article]. Communication Quarterly, 55(4), 451-466.
Berlo, D. K., Lemert, J. B., & Mertz, R. J. (1969). Dimensions for evaluating the acceptability of
message sources. Public Opinion Quarterly, 33(4), 563-576.
Breakenridge, D. (2008). PR 2.0.: New media, new tools, new audiences. Upper Saddle River,
NJ: Pearson Education.
Burgoon, J. K., & Hale, J. L. (1984). The fundamental topoi of relational communication.
Communication Monographs, 51(3), 193.
Byrne, B. M. (2001). Structural equation modeling with AMOS: Basic concepts, applications,
and programming. Mahwah, NJ: Lawrence Erlbaum Associates.
Cutlip, S. M., Center, A. H., & Broom, G. M. (2000). Effective public relations (8th ed.). Upper
Saddle River: NJ: Prentice-Hall.
Edelman, R. (2008, October 30). Public engagement: The evolution of public relations Retrieved
October 30, 2008, from
http://www.edelman.com/speak_up/blog/UMD_presentation/UMD_files/frame.htm
24
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Flanagin, A. J., & Metzger, M. J. (2000). PERCEPTIONS OF INTERNET INFORMATION
CREDIBILITY. [Article]. Journalism & Mass Communication Quarterly, 77(3), 515-540.
Flanagin, A. J., & Metzger, M. J. (2003). The perceived credibility of personal Web page
information as influenced by the sex of the source. [doi: DOI: 10.1016/S0747-
5632(03)00021-9]. Computers in Human Behavior, 19(6), 683-701.
Glaser, B. G., & Strauss, A. L. (1967). The Discovery of Grounded Theory: Strategies for
Qualitative Research: Aldine.
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis.
Upper Saddle River, NJ: Prentice Hall.
Holtz, S., & Havens, J. C. (2009). Tactical Transparency: How leaders can leverage social
media to maximize value and build their brand. San Francisco, CA: Jossey-Bass.
Hostway (2005). Consumer Hostway blog survey: Top-line results. Retrieved September 14,
2010, from http://www.hostway.com/media/survey/blog.html
Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953). Communication and persuasion;
psychological studies of opinion change. New Haven,: Yale University Press.
Hu, L.-t., & Bentler, P. M. (1999). Cutoff Criteria for Fit Indexes in Covariance Structure
Analysis: Conventional Criteria versus New Alternatives. Structural Equation Modeling:
A Multidisciplinary Journal, 6(1), 1-55.
Johnson, T. J., & Kaye, B. K. (1998). CRUISING IS BELIEVING?: COMPARING INTERNET
AND TRADITIONAL SOURCES ON MEDIA CREDIBILITY MEASURES. [Article].
Journalism & Mass Communication Quarterly, 75(2), 325-340.
25
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Johnson, T. J., & Kaye, B. K. (2004). For Whom the Web Toils: How Internet Experience
Predicts Web Reliance and Credibility. [Article]. Atlantic Journal of Communication,
12(1), 19-45.
Johnson, T. J., & Kaye, B. K. (2009). In blog we trust? Deciphering credibility of components of
the internet among politically interested internet users. [Article]. Computers in Human
Behavior, 25, 175-182.
Kiousis, S. (2001). Public trust or mistrust? Perceptions of media credibility in the information
age. Mass Communication and Society, 4(4), 381-403.
Kline, R. B. (1998). Principles and practice of structural equation modeling. New York: The
Guilford Press.
Lenhart, A., & Fox, S. (2006). Bloggers. A portrait of the internet’s new storytellers. Pew
Internet & American Life Project. Retrieved March 29, 2008, from
http://www.pewinternet.org/pdfs/PIP%20Bloggers%20Report%20July%2019%202006.p
df.
McClure, J. (2007, February 26). New tools, new rules, & a new role for PR. Unpublished public
lecture. S. I. Newhouse School of Public Communications, Syracuse University.
McCroskey, J. C., & Richmond, V. P. (1996). Fundamentals of human communication. Prospect
Heights, IL: Waveland Press.
Metzger, M. J., Flanagin, A. J., Eyal, K., Lemus, D., & McCann, R. (2003). Credibility for the
21st century: Integrating perspectives on source, message, and media credibility in the
contemporary media environment. Communication Yearbook, 27, 293 - 335.
Meyer, P. (1988). Defining and measuring credibility of newspapers: Developing an index.
Journalism Quarterly, 65, 567-588.
26
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Newhagen, J., & Nass, C. (1989). Differential criteria for evaluating credibility of newspapers
and TV news. Journalism Quarterly, 66(277-284).
PR News (2009, August 17). 2009 Digital Readiness Report.
Rowse, D. (2006, April 24). Blog credibility. Retrieved September 15, 2010, from
http://www.problogger.net/archives/2006/04/24/blogger-credibility/
Rubin, A. (1984). Ritualized and instrumental television viewing. Journal of Communication,
34(3), 67-77.
Rubin, A. (2002). The uses-and-gratifications perspective of media effects. In J. Bryant & D.
Zillmann (Eds.), Media effects: Advances in theory and research (2nd ed., pp. 525-548).
Mahwah, NJ: Lawrence Erlbaum.
Scoble, R., & Israel, S. (2006). Naked conversations: How blogs are changing the way business
talk with customers. Hoboken, NJ: John Wiley & Sons.
Scott, D. M. (2007). The new rules of Marketing and PR: How to create a press release strategy
for reaching buyers directly. Hoboken, NJ: John Wiley & Sons.
Sundar, S. S. (2008). The MAIN Model: A Heuristic Approach to Understanding Technology
Effects on Credibility. The John D. and Catherine T. MacArthur Foundation Series on
Digital Media and Learning, 73-100.
Sundar, S. S., & Nass, C. (2001). Conceptualizing sources in online news. [Article]. Journal of
Communication, 51(1), 52.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). Needham
Heights, MA: Allyn and Bacon.
27
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Thorson, K., Vraga, E., & Ekdale, B. (2010). Credibility in Context: How Uncivil Online
Commentary Affects News Credibility. [Article]. Mass Communication & Society, 13(3),
289-313.
Weil, D. (2006). The corporate blogging book: Absolutely everything you need to know to get it
right. New York: PORTFOLIO.
Yang, S.-U., Kang, M., & Johnson, P. (2010). Effects of narratives, openness to dialogic
communication, and credibility on engagement in crisis communication through
organizational blogs. Communication Research, 37(4), 473-497.
Yang, S.-U., & Lim, J. (2009). The effects of blog-mediated public relations on relation trust.
Journal of Public Relations Research, 21(3), 341-359.
28
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Table 1. Exploratory Factor Analysis for Measurement Items on Blog Credibility
Factor Loading
Item
Blog Content
Credibility
Blogger
Credibility
Variance
Extracted %
α
1. Knowledgeable .26 .73 24.00 % .84
2. Influential .23 .77
3. Passionate .18 .68
4. Transparent .12 .78
5. Reliable .23 .78
5. Authentic .77 .10 40.24 % .93
6. Insightful .81 .15
7. Informative .80 .15
8. Consistent .80 .23
9. Fair .79 .28
10. Focused .79 .20
11. Accurate .73 .39
12. Timely .74 .32
13. Popular .74 .25
Total Variance
Explained
64.23%
Note. Extraction method was Principal Component Analysis with Varimax Rotation.
29
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Table 2. Correlation Matrix
1 2 3 4 5 6 7 8 9 10 11 12 13 14
1. Knowledgeable
2. Influential .58
3. Passionate .37 .51
4. Transparent .51 .51 .45
5. Reliable .56 .55 .50 .54
6. Authentic .28 .31 .28 .16 .22
7. Insightful .34 .27 .29 .26 .30 .67
8. Informative .34 .37 .24 .21 .28 .65 .64
9. Consistent .39 .34 .30 .29 .36 .55 .63 .62
10. Fair .38 .39 .26 .32 .44 .51 .62 .57 .76
11. Focused .37 .30 .29 .21 .37 .50 .61 .58 .70 .70
12. Accurate .50 .41 .32 .44 .44 .53 .63 .65 .63 .67 .61
13. Timely .38 .42 .31 .37 .39 .56 .57 .64 .58 .63 .62 .67
14. Popular .27 .38 .34 .30 .40 .57 .58 .55 .59 .61 .61 .58 .61
Note. n = 250. All coefficients are statistically significant at the .01 level (2-tailed).
30
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Figure 1. Factor Plot of Blog Credibility Measurement Items.
Note. Extraction method was Principal Component Analysis with Varimax Rotation.
31
Measuring Social Media Credibility – A Study on a Measure of Blog Credibility
by Minjeong Kang
Copyright © 2010, Institute for Public Relations
www.instituteforpr.org
Figure 2. The Final Confirmatory Factor Analysis (CFA) of Blog Credibility Measurement
Items. Note: 2
= 113.77, df = 73, p < .001, χ2
/df = 1.56, CFI = .973, RMSEA = .056 (90%
CI: .035, .075), and SRMR = .042.